Faculty Feature: Scott Moisik
Constellations, Issue 5
Assistant Professor Scott Moisik from the Linguistics and Multilingual programme focuses on the ‘meat’ of speech. He researches speech production with an emphasis on anatomy, physiology, and biomechanics of speech, as well as the genetic, developmental, and evolutionary underpinnings of the vocal tract. He is interested in how these factors, along with the aerodynamic, acoustic, perceptual, and social facets of speech, conspire to shape speech sound systems, both giving rise to striking similarities yet also diversity from one language to the next.
Does the shape of our teeth, lips, tongue, or other parts of our vocal tract influence how we talk? Could languages be similarly shaped by the biology of their speakers (or perhaps even vice versa)? Many people have wondered about such questions, but only recently have we had the data and computing power to take a serious look at them. Several papers co-authored by Prof. Moisik have examined various cases of possible anatomical biases in speech sounds.
The first of these (Moisik & Dediu 2017; covered in Scientific American), published in the Journal of Language Evolution, looks at click sounds, famously found in southern Africa. Using x-ray and other imaging techniques, previous researchers had noticed that many click-language speakers lack a prominent alveolar ridge — a bulge most people have behind their upper front teeth. Prof. Moisik’s paper used a computational biomechanical model of click production to show that lacking this bulge might actually make it easier to produce click sounds, possibly helping us to understand why these sounds are so rare as phonemes (the building blocks of a language’s speech sound system).
The second paper (Blasi, Moran, Moisik, et al. 2019; summarized nicely on SciShow), published in Science, looks at the possibility that ‘f’ and ‘v’ sounds (known as labiodentals) might have become much more widespread across languages thanks to the development of agriculture and subsequent food processing techniques. It turns out that processing food reduces the amount of wear and tear on our teeth, making the upper incisors overlap the lower ones in an ‘overbite’, instead of coming together in what is known as an ‘edge bite’. With the help of biomechanical modelling and a variety of statistical techniques, the paper presents evidence that the reshaping of the jaw caused by an agricultural diet also makes it easier to produce labiodentals, and this is reflected in the current distribution of these sounds among the world’s languages, being far less common in languages associated with hunter-gather populations.
Finally, the last of the papers (Dediu, Janssen, & Moisik 2019), published in Nature Human Behaviour, uses computer-simulated speakers (‘agents’) equipped with realistic vocal tract models to show that palate shape can influence vowel systems across many generations of language transmission. In complement to this, Prof. Moisik shows, using a large sample of 3D hard palate scans collected from real people, that speakers from diverse populations can be differentiated by their palate shape, lending support to the possibility that the results of his computer model apply to real speakers.
Apart from opening a new line of inquiry into the forces that shape language, all these papers together suggest that we cannot divorce language from the biology of its speakers and that we should recognize and embrace our diversity in order to better understand one of the most uniquely human of all our attributes, language.